Catalytic deconstruction of waste polyethylene with ethylene to form propylene

Abstract

The conversion of polyolefins to monomers would create a valuable carbon feedstock from the largest fraction of waste plastic. However, breakdown of the main chains in these polymers requires the cleavage of carbon-carbon bonds that tend to resist selective chemical transformations. Here, we report the production of propylene by partial dehydrogenation of polyethylene and tandem isomerizing ethenolysis of the desaturated chain. Dehydrogenation of high-density polyethylene with either an iridium-pincer complex or platinum/zinc supported on silica as catalysts yielded dehydrogenated material containing up to 3.2% internal olefins; the combination of a second-generation Hoveyda-Grubbs metathesis catalyst and [PdP(^tBu)₃(μ-Br)]₂ as an isomerization catalyst selectively degraded this unsaturated polymer to propylene in yields exceeding 80%. These results show promise for the application of mild catalysis to deconstruct otherwise stable polyolefins.

Fig. 1.. Strategies for the valorization of PE waste.

(A) Degradation of PE by alkane metathesis with n-hexane. (B) Conversion of PE to alkyl aromatics by tandem hydrogenolysis/aromatization. (C) Degradation of PE by hydrogenolysis. (D) Conversion of PE to propylene by DIE (where x is the number of I/E turnovers). Mes, mesitylene (2,4,6-trimethylbenzene).

Fig. 2.. Conditions tested for DIE.

(A) Homogeneous dehydrogenation of PE: 0.42 mol % Ir-^tBuPOCOP (9.8 wt %), 0.46 mol % NaOtBu, 0.6 M in PE; reactions were run for 12 hours with 0.4 equivalents of TBE. Mol %, equivalents. TBE and M of PE were calculated relative to PE monomer units. (B) HDPE (M_n = 26.1 kDa) with bimetallic catalysts supported on alumina. Conditions: 20 wt % catalyst, 350°C. Catalysts were first activated under H₂ flow for 1 hour, and then dehydrogenation was conducted for 16 hours under Ar flow. (C) Gel-permeation chromatogram of HDPE before and after ethenolysis with M1. Conditions for ethenolysis were 0.36 M in unsaturated PE 3.6 mol % M1 in p-xylene (relative to PE monomer units), heated at 130°C for 16 hours under 25 bar ethylene. (D) Sequential dehydrogenation and ethenolysis of a long-chain paraffin (tetracosane). Initial dehydrogenation yields olefins in 10% yield, and subsequent ethenolysis with M1 yields a statistical distribution of shorter-chain olefins.

Fig. 3.. Development of conditions for DIE of PE.

(A) Application of I/E to dehydro-PE, with dehydro-HDPE yielding a maximum of 80% propylene with 1.96% dehydrogenation. Conditions for I/E of dehydrogenated PE: 3.6% M1, 2.2% I1, 1.78 M in PE, heated at 130°C in p-xylene for 16 hours under 25 bar of ethylene. Catalyst loadings are reported as mol % relative to total PE repeat units. (B) DIE of ¹³C-labeled HDPE to check for background reactivity [conditions identical to those in (A)]. (C) DIE of postconsumer PE [conditions identical to those in (B)]. Propylene yields are reported as X% (Y) where X is the yield obtained from polymer purified by precipitation from acetone and Y is the yield obtained from unmodified polymer. Where a single yield is reported, dehydrogenation on unpurified samples failed.

Fig. 4.. Investigation of reaction scope and kinetics.

(A) Time course of I/E with 1-octadecene (100% yield = 458.3 mbar). Conditions for I/E of 1-octadecene: 6.0 mol % M1 (0.67 mol % relative to ethylene subunits), 3.0 mol % I1 (0.33 mol % relative to ethylene subunits), 0.500 M in olefin, heated at 60°C in p-xylene for 16 hours under 25 bar of ethylene. A fit for the time course was conducted using COPASI by simulating an exponential appearance of catalyst and a zero-order reaction (R² = 0.993, k_cat = 0.071 s⁻¹, k_propylene = 18.0 s⁻¹). (B) Time course of I/E with dehydrogenated HDPE (1.41% olefin) (100% yield = 193.5 mbar). A linear fit has been added to (B) to show the zero-order reaction after the short induction period and completion of the reaction (R² = 0.989). (C) Yields of I/E on small molecules 1a to 1d designed to assess the effect of branching on I/E. (D) Yields of I/E on small molecules 1e and 1f designed to test the effect of dienes on I/E [conditions identical to those specified in (A)] and on PCO (M_n= 5.2 kDa) designed to demonstrate the effect of polymer unsaturation on I/E yield in the absence of branching. Conditions for I/E in (C) and (D) were identical to those in Fig. 3C.